KR20070001872A - System and method for using a simulation of the first principles in a semiconductor manufacturing process - Google Patents

Abstract

A method, system, and computer readable medium for facilitating a process performed by a semiconductor processing tool. The method includes entering data related to a process performed by a semiconductor processing tool and entering a physical model of first principles related to the semiconductor processing tool. Simulation of the first principles is performed using input data and a physical model to provide virtual sensor measurements related to the process performed by the semiconductor processing tool, and the virtual sensor measurements are used to facilitate the process performed by the semiconductor processing tool. . In addition, simulations of the first principles are performed using input data and a physical model to provide simulation results for the process performed by the semiconductor processing tool, and then the data set characterizing the process where the simulation results are performed by the semiconductor processing tool. Also disclosed is a method used as part of.

Description

SYSTEM AND METHOD FOR USING FIRST-PRINCIPLES SIMULATION IN A SEMICONDUCTOR MANUFACTURING PROCESS

FIELD OF THE INVENTION The present invention relates generally to manufacturing semiconductor devices and, more particularly, to using first principles simulation in semiconductor manufacturing processes.

Material processing in the semiconductor industry presents a formidable challenge in the manufacture of integrated circuits. The demands to increase the speed of common ICs, and in particular memory devices, force semiconductor manufacturers to make devices on the substrate surface smaller. In addition, to reduce fabrication costs, the overall complexity of the IC structure and its processing methods can be reduced by reducing the number of steps (e.g., etching steps, deposition steps, etc.) needed to generate the IC structure. Should be reduced. These requirements increase substrate size (ie, 200 mm to 300 mm), which emphasizes the reduction of feature size and the precise control of critical dimensions, process speed, and process uniformity to maximize yield of advanced devices. Above) is aggravated by all.

In semiconductor fabrication, many steps are used during the evolution of ICs, including vacuum processing, thermal processing, plasma processing, and the like. Within each processing step, there are a number of variables that affect the outcome of the processing. In order to more precisely control the results of each processing step, the individual processing tools are increasingly equipped with (electrical, mechanical and optical) diagnostic systems for measuring data during processing, through the actions of the process controller. It provides an intelligent basis for correcting variations. Many diagnostic systems are increasing burdens and costs. However, data that is decisive enough in space and time for complete process control is still not available.

Due to these industry and manufacturing difficulties, the semiconductor manufacturing industry has drawn a lot of attention to the use of computer-based modeling and simulation. Computer-based modeling and simulation is widely used to predict tool performance during the design process of semiconductor manufacturing tools. The use of modeling enables the reduction of both cost and time associated with the tool development cycle. Modeling in many areas, such as stress, heat, magnetism, etc., has reached a maturity level where modeling can be trusted to provide accurate answers to design questions. In addition, computer capabilities have increased rapidly with the development of new solution algorithms, all of which have reduced the time required to obtain simulation results. Indeed, the inventors have found that the majority of simulations that are typically made at the time of tool design can now be executed at times comparable to wafer or wafer cassette processing times. These trends have led to the suggestion that simulation functions, which are typically only used for tool design, can be implemented directly in the tool itself, contributing to the various processes performed by the tool. For example, in 2001, the International Technology Roadmap for Semiconductors (ITRS), which issues issues that delay the development of on-tool integrated simulation capability, produces very small features in future semiconductor devices. I agreed that this is a technology that makes things possible.

Indeed, the industry's failure to implement on-tool simulations to facilitate tool processes is largely due to the need for computational resources that can perform simulations in a reasonable time. Specifically, processor functions currently dedicated to semiconductor fabrication tools are typically limited to diagnosing and controlling functions, so that only relatively simple simulations can be performed. Therefore, the semiconductor manufacturing industry recognized that to realize meaningful on-tool simulation functions, it was necessary to provide powerful dedicated computers. However, such computer-only to semiconductor processing tools creates wasted computational resources when the tool runs processes that use simple simulations or do not use simulations at all. This inefficient use of expensive computational resources has been a major stumbling block for implementing simulation functions in semiconductor processing tools.

One object of the present invention is to reduce or solve the above identified and / or other problems with the prior art.

Another object of the present invention is to integrate the simulation functions of the first principles with a semiconductor fabrication tool to facilitate the process performed by the tool.

Yet another object of the present invention is to provide tool simulation functions without the need for powerful computational resources dedicated to the tool.

It is yet another object of the present invention to provide a wide range of on-tool simulation functions using existing computational resources dedicated to each tool in a manufacturing facility.

These and / or other objects may be provided by the following aspects of the invention.

According to one aspect of the invention, a method for facilitating a process performed by a semiconductor processing tool comprises the steps of inputting data related to a process performed by the semiconductor processing tool and a physical model of first principles relating to the semiconductor processing tool. principles physical model). Simulation of the first principles is performed using input data and a physical model to provide virtual sensor measurements related to the process performed by the semiconductor processing tool, and the virtual sensor measurements facilitate the process performed by the semiconductor processing tool. It is used to

According to another aspect of the invention, a system includes a semiconductor processing tool configured to perform a process and an input device configured to input data related to a process performed by the semiconductor processing tool. The simulation processor of the first principles comprises a virtual sensor associated with a process performed by the semiconductor processing tool to input a physical model of the first principles related to the semiconductor processing tool and to perform simulations of the first principles using the input data and the physical model. It is configured to provide a measurement. Virtual sensor measurements are used to facilitate the process performed by the semiconductor processing tool.

According to another aspect of the invention, a system for facilitating a process performed by a semiconductor processing tool comprises a means for inputting data related to a process performed by the semiconductor processing tool and a physical of first principles relating to the semiconductor processing tool. Means for inputting the model. Means for performing simulations of the first principles using input data and physical models to provide virtual sensor measurements related to the process performed by the semiconductor processing tool and facilitating the process performed by the semiconductor processing tool using the virtual sensor measurements. Means for doing so are also included.

In another aspect of the invention, when executed by a computer system, the processor performs a step of inputting data relating to a process performed by a semiconductor processing tool and entering a physical model of first principles related to the semiconductor processing tool. A computer readable medium is provided that includes program instructions for executing on a processor. The processor also performs simulations of the first principles using input data and physical models to provide virtual sensor measurements related to the process performed by the semiconductor processing tool, and facilitates the process performed by the semiconductor processing tool. To use virtual sensor measurements.

Another aspect of the invention is a method of facilitating a process performed by a semiconductor processing tool, the method comprising the steps of inputting data related to the process performed by the semiconductor processing tool and of the first principles related to the semiconductor processing tool. Inputting a physical model. Next, a simulation of the first principles is performed using input data and a physical model to provide simulation results for the process performed by the semiconductor processing tool, which simulation data characterizes the process performed by the semiconductor processing tool. Used as part of a set.

Another aspect of the invention is a system comprising a semiconductor processing tool configured to perform a process and an input device configured to input data related to a process performed by the semiconductor processing tool. The simulation processor of the first principles uses the input data and the physical model to input a physical model of the first principles related to the semiconductor processing tool and to provide a simulation result of the first principles for the process performed by the semiconductor processing tool. Configured to perform a simulation of the first principles. The simulation results are used as part of the data set characterizing the process performed by the semiconductor processing tool.

Another aspect of the invention is a system for facilitating a process performed by a semiconductor processing tool, the means for inputting data relating to a process performed by the semiconductor processing tool and a physical model of first principles related to the semiconductor processing tool. It is a system comprising means for inputting. The system features means for performing simulations of the first principles using input data and physical models to provide simulation results for processes performed by semiconductor processing tools and processes performed by semiconductor processing tools for simulation results. It also includes a means for use as part of the data set.

Another aspect of the invention, when executed by a computer system, causes a processor to enter data relating to a process performed by a semiconductor processing tool and to input a physical model of first principles related to the semiconductor processing tool. Computer-readable medium containing program instructions for execution on a processor. In addition, the processor may perform simulations of the first principles using input data and physical models to provide simulation results for the processes performed by the semiconductor processing tool, and characterize the processes performed by the semiconductor processing tool. To be used as part of the data set.

By reference to the following detailed description considered in conjunction with the accompanying drawings, a more complete understanding of the present invention and many of the advantages accompanying it will be readily obtained.

1 is a block diagram of a system for facilitating a process performed by a semiconductor processing tool using simulation techniques of the first principles, in accordance with an embodiment of the present invention.

2 is a flow diagram illustrating a process for facilitating a process performed by a semiconductor processing tool using simulation techniques of the first principles, in accordance with an embodiment of the present invention.

3 is a block diagram of a network architecture that may be used to provide simulation techniques of the first principles to facilitate a process performed by a semiconductor processing tool, in accordance with an embodiment of the invention.

4 is a block diagram of a system for providing virtual sensor measurements for a semiconductor processing tool using simulation techniques of the first principles, in accordance with an embodiment of the present invention.

5 is a block diagram of a system for characterizing a process on a semiconductor processing tool using simulation techniques of the first principles, in accordance with an embodiment of the present invention.

6 is a block diagram of a system for controlling a process performed by a semiconductor processing tool using simulation techniques of the first principles, in accordance with an embodiment of the present invention.

7 is a flowchart illustrating a process for controlling a process performed by a semiconductor processing tool using simulation techniques of the first principles, in accordance with an embodiment of the present invention.

8 is a block diagram of a system for controlling a process performed by a semiconductor processing tool using simulation techniques and empirical models of the first principles, in accordance with an embodiment of the present invention.

9 is a flowchart illustrating a process for controlling a process performed by a semiconductor processing tool using simulation techniques and empirical models of first principles, in accordance with an embodiment of the present invention.

10 is a block diagram of a system for controlling a process performed by a semiconductor processing tool using an error detector and simulation techniques of the first principles, in accordance with an embodiment of the present invention.

및

) 및 대응되는 출력들(

,

,

,

,

,

,

,

)과 VIP(variable importance in the projection)의 개략적인 표현이다. 11 shows data inputs for PLS analysis (

And

) And corresponding outputs (

,

,

,

,

,

,

,

) And VIP (variable importance in the projection).

12 is a flowchart illustrating a process for detecting errors and controlling processes performed by a semiconductor processing tool using simulation techniques of the first principles, in accordance with an embodiment of the present invention.

13 is a block diagram of a vacuum processing system to which a process control embodiment of the present invention may be applied.

14 illustrates a computer system in which embodiments of the present invention may be implemented.

(Description of Preferred Embodiments)

Referring now to the drawings, wherein like reference numerals designate the same or corresponding parts throughout the several views, FIG. 1 illustrates a semiconductor processing tool using simulation techniques of the first principles, in accordance with an embodiment of the invention. Is a block diagram of a system to facilitate the process performed by a. As can be seen in FIG. 1, the system includes a semiconductor processing tool 102, a data input device 104, a physical model 106 of the first principles, and a simulation processor 108 of the first principles. The system of FIG. 1 may also include a tool level library 110 as shown by phantomization.

The semiconductor processing tool 102 is a tool for performing a process related to fabricating an integrated circuit or semiconductor wafer. For example, the semiconductor processing tool 102 may be a material processing system, an etching system, a photoresist spin coating system, a lithography system, an insulating coating system (i.e., spin-on-glass (SOG) or SOD). (spin-on-dielectric) systems, deposition systems (i.e. chemical vapor deposition (CVD) or physical vapor deposition (PVD) systems), rapid thermal processing (RTP) systems for thermal annealing, batch diffusion furnace), or any other tool for a semiconductor manufacturing process.

The data input device 104 is a device for collecting data related to the process performed by the semiconductor processing tool 102 and inputting the collected data into the simulation processor 106 of the first principles. The process performed by the semiconductor process tool 102 may be a characterization process (ie, process design or development), a cleaning process, a production process, or any other process performed by the semiconductor processing tool. In one embodiment, the data input device 104 may be implemented as a physical sensor for collecting data related to the semiconductor processing tool 102 itself and / or the environment contained within the chamber of the tool. Such data may include mechanical data such as gas velocities and pressures at various locations within the process chamber, electrical data such as voltage, current, and impedance at various locations within the electrical system of the process chamber, various data within the process chamber. Chemical data such as species concentrations and reaction actions at locations, thermal data such as gas temperature, surface temperature, and surface heat flux at various locations in the process chamber, if plasma is used Plasma processing data such as plasma density (eg, obtained from Langmuir probes), ion energy (eg, obtained from an ion energy spectrum analyzer), and pressure, deflection, stress, at various locations within the process chamber, And mechanical data such as strain.

In addition to the tool and tool environment data, the data input device 104 may collect data related to the process itself or process results obtained on the semiconductor wafer on which the tool 102 is performing the process. In one embodiment, the data input device 104 is implemented as a metrology tool coupled to the semiconductor processing tool 102. The metrology tool may include etch rate, deposition rate, etch selectivity (ratio of the rate at which the first material is etched relative to the rate at which the second material is etched), an etch critical dimension (eg, length or width of the feature). ), Anisotropy of etching features (eg, sidewall profiles of etching features), film properties (eg, film stress, porosity, etc.), film thickness of masks (eg, photosensitive resin films), masks (eg For example, it may be configured to measure process performance parameters such as the pattern critical dimension of the photosensitive resin film (or any other parameter for the process performed by the semiconductor processing tool 102).

The data input device is coupled directly to the process tool 102 and the simulation processor 106 of the first principles, as shown in FIG. 1, to automatically receive data from the tool 102 and to transfer the data to the first principle. To the simulation processor 106. Alternatively, the data input device 104 may be implemented as user input data used to indirectly provide the simulation processor 106 with data related to the process performed by the semiconductor processing tool 102. For example, the data input device 104 may be a keyboard that a simulation operator uses to enter data into the simulation processor 106 of the first principles. Alternatively, the data input device may be a database for storing data related to processes performed in the past by the semiconductor processing tool 102. In this embodiment, the database may be populated automatically and / or by manual input by the use of a physical sensor or metrology tool coupled to the semiconductor processing tool 102. The database may be automatically accessed by the simulation processor 108 of the first principles to enter data into the processor.

The physical model 106 of the first principles is not only physical to the tools and the tool environment, but also to the basic equations required to perform the simulation of the first principles and provide simulation results to facilitate the process performed by the semiconductor processing tool. It is a model of attributes. Thus, the physical model 106 of the first principles depends to some extent not only on the type of semiconductor processing tool 102 being analyzed but also on the process performed on the tool. For example, the physical model 106 can include a spatially resolved model of the physical geometry of the tool, for example, different for chemical vapor deposition (CVD) chambers and diffusion paths. Likewise, the equations of the first principles necessary to calculate flow fields are quite different from the equations needed to calculate temperature fields. The physical model 106 is used to calculate ANSYS Inc., in order to calculate the flow paths, the magnetic fields, the temperature fields, the chemical properties, the chemical properties of the surface (ie, the chemical properties of the etching surface or the chemical properties of the deposition surface). Southpointe, 275 Technology Drive Canonsburg, PA 15317, ANSYS, Fluent Inc., 10 Cavendish Ct. It may be a model implemented with commercial software, such as FLUENT from Centerra Park, Lebanon, NH 03766, or CFD-ACE + from CFD Research Corp., 215 Wynn Dr., Huntsville, AL 35805. However, special purpose or custom models developed from the first principles may be used to address these and other details in the processing system.

The simulation processor 108 of the first principles is a processing device that applies data input from the data input device 104 to the physical model 108 of the first principles to perform a simulation of the first principles. Specifically, the simulation processor 108 of the first principles may use the data provided by the data input device 104 to set initial conditions and / or boundary conditions for the physical model 106 of the first principles, The physical model 106 of the first principles is then executed by the simulation module. Simulations of the first principles in the present invention are continuum for mass, momentum, and energy transport derived from simulations of electro-magnetic fields derived from Maxwell's equations, continuous Navier-Stokes equations, and the first law of thermodynamics. Not only simulations, but also atomic simulations derived from Boltzmann equations, such as, for example, Monte Carlo simulations of lean gases (Bird, GA 1994. Molecular gas dynamics and the direct simulation of gas flows, Clarendon Press), It is not limited to this. The simulation processor 108 of the first principles may be implemented as a processor or workstation physically integrated with the semiconductor processing tool 102, or as a general purpose computer system, such as the computer system 1401 of FIG. 14. The output of the simulation processor 108 of the first principles is a simulation result used to facilitate the process performed by the semiconductor processing tool 102. For example, simulation results can be used to provide virtual sensor outputs that facilitate tool processes as well as facilitating process development, process control and error detection, as discussed further below.

As shown by phantomization in FIG. 1, the system may include a tool-level library 108 for storage of simulation results. The library is actually a compilation of past simulation results that can be used to provide future simulation results. The tool level library 110 may be stored in a separate storage device or in a computer storage device, such as a hard disk, integrated with the simulation processor 106 of the first principles.

As many variations of the specific hardware and software used to implement the invention will be apparent to those skilled in the art, it will be understood that the system of FIG. 1 is for illustrative purposes only. For example, the functionality of the physical model 106 of the first principles, the simulation processor 108 of the first principles, and the tool level library 110 may be combined in a single device. Likewise, the functionality of the data input device 104 may be combined with the functionality of the semiconductor processing tool 102 and / or the simulation processor 108 of the first principles. To implement these as well as other variations, a single computer (eg, computer system 1401 of FIG. 14) may be programmed to perform two or more special purpose functions of the devices shown in FIG. 1. On the other hand, two or more programmed computers may replace one of the devices shown in FIG. For example, to increase the robustness and performance of the system, the principles and advantages of distributed processing, such as redundancy and replication, may be implemented if desired.

2 is a flow diagram illustrating a process for facilitating a process performed by a semiconductor processing tool using simulation techniques of the first principles, in accordance with an embodiment of the present invention. The process shown in FIG. 2 may be executed, for example, in the simulation processor 104 of the first principles of FIG. 1. As shown in FIG. 2, the process begins with step 201 of entering data related to a process performed by the semiconductor processing tool 102. As discussed above, the input data may be data related to the physical attributes of the tool / tool environment and / or data related to the process performed on the semiconductor wafer by the tool or the results of such a process. In addition, as described above, input data may be input directly from a physical sensor or metrology tool coupled to the simulation processor 104 of the first principles or indirectly from a manual input device or database. When data is indirectly input from a manual input device or database, the data may be data recorded from a previously executed process, such as sensor data from a previously executed process. Alternatively, the data may be set by the simulation operator as "best known input parameters" for certain simulations, which may or may not be related to the data collected during the process. The type of input data input by the processing tool generally depends on the desired simulation result.

In addition to entering the input data, the simulation processor 104 of the first principles also inputs a physical model 106 of the first principles, as indicated by step 203. Step 203 is a first principle, organized in software, that is required to perform simulation of the first principles for certain attributes of the process performed by the semiconductor processing tool 102 as well as the physical attributes of the tool modeled by the model. Inputting the equations thereof. The physical model 106 of the first principles may be input to the processor from an external memory or an embedded memory device integrated into the processor. In addition, although step 203 is shown in FIG. 2 as accompanying step 201, it should be appreciated that the simulation processor 104 of the first principles may perform these steps simultaneously or in reverse order shown in FIG. 2.

In step 205, the simulation processor 108 of the first principles executes the simulation of the first principles using the input data of step 201 and the physical model of the first principles of step 203 and provides simulation results. Step 205 may be performed concurrently or asynchronously with the process performed by the semiconductor processing tool. For example, simulations that can be performed in short solution times can be run concurrently with the tooling process, and the results are used to control the process. More computationally intensive simulations can be performed asynchronously with the tooling process, and simulation results can be stored in a library for later retrieval. In one embodiment, step 205 includes setting initial and / or boundary conditions for the physical model provided in step 205 using the input data of step 201.

Once the simulation is run, the simulation results are used to facilitate the process performed by the semiconductor processing tool 102. As used herein, the term "facilitates a process performed by a semiconductor processing tool" refers to, for example, detecting an error in a process, controlling a process, or characterizing a process for manufacturing operations. Or use the simulation result to provide virtual sensor readings related to the process, or any other use of the simulation result in connection with facilitating a process performed by the semiconductor processing tool 102.

3 is a block diagram of a network architecture that may be used to provide simulation techniques of the first principles to facilitate a process performed by a semiconductor processing tool, in accordance with an embodiment of the invention. As can be seen in this figure, the network architecture includes a device fabrication Fab that is connected to remote resources via the Internet 314. The device fabrication Fab includes a plurality of semiconductor processing tools 102 that are connected to individual simulation modules 302. As described with reference to FIG. 1, each semiconductor processing tool 102 is a tool for performing a process related to manufacturing a semiconductor device, such as an integrated circuit. Each simulation module 302 is a computer, workstation, or other processing device capable of implementing the simulation techniques of the first principles to facilitate the process performed by the semiconductor processing tool 102. Thus, each simulation module 302 may be helpful in executing simulations of the first principles, as well as the physical model 106 of the first principles and the simulation processor 106 of the first principles described with reference to FIG. 1. Any other hardware and / or software that may be. In addition, the simulation modules 302 may be configured to communicate with an Fab-level advanced process control (APC) controller using any known network communication protocol. Each simulation module 302 may be implemented as a general purpose computer system, such as computer system 1401 of FIG. 14.

Although not shown in FIG. 3, each simulation module 302 is associated with a data input device for entering data related to the process performed by the tool 102. In the embodiment of FIG. 3, the simulation modules 302 are directly coupled to the individual tool 102, so that the data input device is physically mounted and / or instrumented on the individual tool 102. Implemented as a tool. However, as described above, the data input device may be implemented as a manual input device or a database used by the simulation module operator. In addition, each simulation module 302 may be configured to store information in and retrieve information from a tool-level library, such as library 306. In addition, as noted above, the tool level library is substantially a compilation of past simulation results that may be useful for future simulations.

In one embodiment of the invention, each simulation module 302 is connected to a main Fab-level APC controller 304 via network connections. As can be seen in FIG. 3, the Fab-level APC controller 304 is a standalone simulation module as well as a standalone simulation module 308 and Fab-level library 310 via the Internet 314 and communication server 316. 312 may also be connected.

Standalone simulation modules 308 and 312 are computational resources that can be used to assist the simulation modules 302 in performing simulations of computationally intensive first principles, as discussed further below. Fab-level library 310 is a database for storing simulation results obtained from any of the simulation modules of the network system. Fab-level APC controller 304 is any workstation, server suitable for communicating with simulation modules 302, 308, and 312 and for storing information in and retrieving information from Fab-level libraries 310. , Or other device. In addition, the Fab-level APC controller 304 facilitates the processes performed by the tools 102 based on the simulation results of the simulation modules 302. For example, the APC controller may be configured to receive simulation results from the simulation module and to use the simulation results to implement a control method for process adjustment and / or correction of any of the tools 102. The Fab-level APC controller 304 communicates with the simulation modules 302, 308 and 312 and the Fab-level library 310 using any suitable protocol, eg, the computer system 1401 of FIG. 14. Can be implemented using

The inventors have found that the configuration of FIG. 3 provides a meaningful on-line that may facilitate the processes performed by the tool by providing computational and storage resource sharing that allows simulation results of a wide range of first principles at moderate solution speeds. It can be seen that it provides tool simulation functions. Specifically, simple simulations can be executed by a dedicated simulation module of the tool, but complex simulations that require more computational resources can be code parallelization techniques in multiple simulation modules of the network, which can be on-tool or standalone. techniques can be used. If power exists for the simulation module, even on-tool simulation modules of the equipment currently under preventive maintenance can be used as a shared computing resource. Likewise, simulation results used for future lookups can be stored in libraries (eg, storage devices) anywhere in the Fab network so that they can be accessed by all tools when lookups of diagnostic or control data are formed. have.

In addition, the inventors have found that the network architecture of FIG. 3 distributes model results performed in one processing tool 102 for one set of conditions to similar or identical other tools that operate according to the same or similar conditions later. It was also found that by providing the capability, unnecessary simulations were eliminated. Running simulations only for unique processing conditions in on-tool and stand-alone modules and re-using results from similar tools with known simulation solutions for diagnostics and control over a wide range of processing conditions. This allows for rapid development of lookup libraries containing results that can be used. In addition, reusing known solutions as initial conditions for the simulation of the first principles reduces computational requirements and facilitates generating simulation solutions in a time frame that does not conflict with on-line control. Similarly, the network architecture of FIG. 3 also provides the ability to propagate changes and refinements made to physical models and model input parameters from one simulation module to others in the network. For example, during process executions and parallel executions of the model, if it is determined that some input parameters should be changed, these changes may be propagated through the network to all other simulation modules and tools.

In addition, the network architecture of FIG. 3 also allows for selective access to remote computational resources including simulation modules that can be helpful in executing simulation tasks and delivering results back to the device manufacturer Fab. The connection to remote resources can be performed using a secure connection, such as a virtual private network (VPN). Such secure connections can also be established for third parties who provide computational resources to support simulation of the first-principles through processing tools. Similarly, telecommunication servers act as a "clearing house" for the latest software, models, input parameters, and simulation results available to multiple customers, further increasing the speed at which accurate result libraries are generated. Can be increased. These updated models are uploaded to the remote resources from the customer site, analyzed, and if it is determined that the improvement applies to the majority of customers, the improvement can be made available to other customers through the communication server and the Internet connection.

As such, the inventors have found meaningful on-tool simulation functions that can facilitate the processes performed by the tool without the need for expensive computers dedicated to the tool. Based on this finding, we additionally provide virtual sensor readings, provide characterization data for use in developing processes performed by the tool, and provide process error detection and process control functions. New on-tool simulation systems have been developed. These uses of the on-tool simulations of the present invention to facilitate the process performed by semiconductor processing tools are by means of one tool and simulation module, or an interconnected network of computational and storage resources as described in FIG. 3. It can be implemented by.

Specifically, on-tool simulation results can be used to magnify data sets measured from physical sensors. One of the disadvantages of current-generation semiconductor processing tools is the relatively small number of sensors used to characterize the processes currently running, especially in production tools. If the number of sensors required is large, installing more sensors in the tool would be a very expensive proposition, and in most cases there is no space left on the tool for modification and installation of additional sensors. Also, even in production tools, there are situations where "measurements" are needed at locations where sensors cannot be installed. The on-tool simulation function of the first principles of the present invention provides the necessary "measurements" without any additional hardware if there are robust models for predicting the measurements using other actual measurements as initial and / or boundary conditions. In this specification, the term “virtual sensor” is used to refer to a “sensor” in which measurements are actually provided by predictions from on-tool simulation.

4 is a flow diagram illustrating a process for providing virtual sensor readings that may facilitate a process performed by a semiconductor processing tool using simulation techniques of the first principles, in accordance with an embodiment of the present invention. The process shown in FIG. 4 may be executed, for example, in the simulation processor 108 of the first principles of FIG. 1, or using the network architecture of FIG. 3. As shown in FIG. 4, the process begins at step 401 of entering data to obtain a virtual sensor reading related to the process performed by the semiconductor processing tool 102. The data input at step 401 can be one of the data types described with reference to step 201 of FIG. 2 if the input data can only cause the simulation of the first principles to provide a virtual sensor simulation result. Thus, the input data may be data related to the physical attributes of the tool / tool environment, the process performed on the semiconductor wafer by the tool, or the results of such a process. In addition, the input data of step 401 may be input directly from a physical sensor or metrology tool coupled to the simulation processor 108 of the first principles, or indirectly from a manual input device or database.

In one example of using metrology data as input data to obtain a virtual sensor reading, the metrology data related to the etch mask pattern and underlying film thickness act as input to the etch process model of the first principles and subsequently performed etch process. can do. Prior to performing the etching process, the mask pattern includes the mask film thickness and the pattern critical dimension (s) at one or more locations (eg, center and edge) on a given substrate for a given substrate lot. The measurements can be provided as input to the etch process model. In addition, measurements of the bottom film thickness (ie, film thickness of the film to be etched) can also act as input to the etching process model. With the execution of the etching process model of the first principles for the specified process method and the metrology input data identified above, for example, the time to complete the etching process at the center and the edge can be calculated as the output, This output can be used, for example, to determine any process adjustments and over-etch periods needed to preserve the center-to-edge of feature threshold dimensions. These results can then be used to adjust the process method for the current or impending substrate lot.

When data is indirectly input by a manual input device or database, the data may be data recorded from a previously executed process, such as sensor data from a previously executed process. Alternatively, the data may be set by the simulation operator as "most famous input parameters" for certain simulations, which may or may not be related to the data collected during the process. The type of input data input by the processing tool generally depends on the predetermined virtual sensor measurements to be obtained.

In addition to entering the input data, the simulation processor 108 of the first principles also inputs a physical model of the first principles for emulating a physical sensor, as indicated by step 403. Step 403 performs a simulation of the first principles to obtain virtual sensor readings that can replace physical sensor readings related to the process performed by semiconductor processing tool 102 as well as physical properties of the tool modeled by the model. Inputting basic equations of the first principles necessary to do so. The physical model of the first principles of step 403 may be input to the processor from an external memory or an internal memory device integrated into the processor. Furthermore, although step 403 is shown in FIG. 4 as accompanying step 401, it should be understood that the simulation processor 104 of the first principles may perform these steps simultaneously or in reverse order shown in FIG. 4. .

In step 405, a simulation processor of the first principles, such as processor 108 of FIG. 1, performs a simulation of the first principles using a physical model of the input data of step 401 and the first principles of step 403 and performs virtual sensor measurements. to provide. Step 405 may be performed at a different time or concurrently with the process performed by the semiconductor processing tool. Simulations that are not executed concurrently with the wafer process may use the initial and boundary conditions stored from preceding process runs with the same or similar process conditions. As noted above with reference to FIG. 2, this is appropriate in cases where the simulation runs slower than the wafer process, where the time between wafer cassettes and even during tool shutdowns, for example, the measurements required by the simulation module. It can be used for preventive maintenance to get them out. These “measurements” can be displayed during subsequent wafer processes, as if they were resolved simultaneously with the wafer process, and as if they were run under the same process conditions as the simulation was run.

If the simulation of the first principles is executed concurrently with the process performed by the semiconductor tool, the data entered in step 401 is physically mounted on the semiconductor processing tool to sense parameters determined during the process executed by the tool. Data from sensors. In this embodiment, steady-state simulations are run repeatedly in parallel with the process by iteratively updating boundary conditions for the simulation model of the first principles using physical sensor measurements. The generated virtual measurement data is useful for monitoring by tool operators and is not at all different from measurements made by physical sensors. However, it is desirable that the simulation be able to run quickly, so that the virtual measurements can be updated at an appropriate rate (eg, "sample rate"). In addition, the simulation of the first principles may be performed concurrently without the use of physical sensor input data. In this embodiment, the initial and boundary conditions for the simulation are set based on the readings of the physical sensors prior to the initial setup and execution of the tool before the tool process, and then during the tool process, but independent of the tool process. In this case, a full time-dependent simulation is performed. The acquired virtual measurements can be displayed to the operator and analyzed by the operator like any other tool parameter actually measured. If the simulation runs faster than the wafer process, the simulation results are known prior to the corresponding actual measurements made during the wafer process. Knowing the measurements in advance in time allows the implementation of various feed-forward control functions based on these measurements, as discussed further below.

In another embodiment of the process of FIG. 4, the simulation of the first principles may be performed in a self correction mode by comparing virtual sensor measurements with corresponding physical sensor measurements. For example, during the first run with a given process method / tool condition, the tool operator will use "the most famous input parameters of the day" for the model. During each simulation and after each simulation is executed, the simulation module (s) can compare the predicted “measures” with the actual measurements at the locations where the actual measurements from the physical sensors are performed. have. If a significant difference is detected, optimization and statistical methods may be used to alter the physical model of the input data and / or the first principles until a better match between the predicted data and the actual measurement data is realized. Depending on the situation, such further refinement simulation runs can be made concurrently with subsequent wafer / wafer cassettes or when the tool is off-line. Once the improved input parameters are known, they can be stored in a library for later use, thereby eliminating the need for subsequent input parameters and model improvements for the same process condition. Further, improvements in model and input data can be distributed to other tools through the network configuration of FIG. 3, thereby eliminating the need for self-correcting practices in those other tools.

Once the simulation is run to provide the virtual sensor measurements, the virtual sensor measurements are used to facilitate the process performed by the semiconductor processing tool 102. For example, virtual sensor measurements can be used for a variety of purposes, such as comparison with actual sensor measurements, method changes in the process, error detection and operator alerts, process conditions, creation of databases for model and input data improvement, and the like. Can be used as inputs to a tool control system. These are typical actions performed by the tool control system based on measurements made by physical sensors. The use of virtual sensor measurements may be used to characterize or control the process, as described below. In addition, the virtual sensor measurements are stored in libraries on computer storage media for future use, so that if there is no change in the model or input conditions (eg, during improvement), the simulation runs need to be repeated with the same input conditions. Can be removed.

In addition to providing virtual sensor readings, the on-tool simulation functionality of the first principles of the present invention facilitates semiconductor process development. More specifically, process characterization in the tool by using the current design for the approach of process development attempts requires different process execution for each change to operating parameters, which is a time consuming and expensive characterization. Results in processes. On-tool simulation functions of the first principles of the present invention allow for parameter changes, which, if analysis of the tool itself, include actual process executions, including changes in those process variables that are well modeled by the simulation of the first principles. Allow parameter changes without This can greatly reduce the number of attempts needed to characterize the tooling process.

5 is a flowchart illustrating a process for characterizing a process performed by a semiconductor processing tool using simulation techniques of the first principles, in accordance with an embodiment of the present invention. The process shown in FIG. 5 may be executed, for example, in the simulation processor 108 of the first principles of FIG. 1, or using the architecture of FIG. 3. As shown in FIG. 5, the processor begins at step 501 of entering data to obtain characterization information related to a process performed by the semiconductor processing tool 102. The data input at step 501 is described with reference to step 201 of FIG. 2 if the input data only allows the simulation of the first principles to provide simulation results used to characterize the process performed by the semiconductor processing tool. It may be one of the data types. Thus, the input data may be data related to the physical attributes of the tool / tool environment, the process performed on the semiconductor wafer by the tool, or the results of such a process. In addition, the input data of step 501 may be input directly from a physical sensor or metrology tool coupled to the simulation processor 104 of the first principles, or indirectly from a manual input device or database. In addition, data may be input from a simulation module that provides virtual sensor readings, as described with reference to FIG. 4. When data is indirectly input by a manual input device or database, the data may be data recorded from a previously executed process, such as sensor data from a previously executed process. Alternatively, the data may be set by the simulation operator as "most famous input parameters" for a particular simulation, which may or may not be related to the data collected during the process. The type of input data input by the processing tool generally depends on the predetermined characterization data to be obtained.

In addition to inputting the input data, the simulation processor 108 of the first principles also inputs a physical model of the first principles to characterize the process, as indicated by step 503. Step 503 is typically organized in software, which is necessary to perform simulations of the first principles to obtain characterization data for the process performed by the semiconductor processing tool 102 as well as the physical properties of the tool modeled by the model. Inputting the basic equations of the first principles. The physical model of the first principles of step 503 may be input into the processor from an external memory or an internal memory device integrated into the processor. Furthermore, although step 503 is shown in FIG. 5 as accompanying step 501, it should be understood that the simulation processor 108 of the first principles may perform these steps simultaneously or in reverse order shown in FIG. 5. .

In step 505, a simulation processor of the first principles, such as processor 108 of FIG. 1, performs a simulation of the first principles and characterizes the process using a physical model of the input data of step 501 and the first principles of step 503. Provide the simulation results used. Step 505 may be performed at a different time or concurrently with the process performed by the semiconductor processing tool. Simulations run asynchronously with the tooling process may use stored initial and boundary conditions from preceding process runs with the same or similar process conditions. As noted above with reference to FIG. 2, this is appropriate in cases where the simulation runs slower than the wafer process, and the time between wafer cassettes and even during tool shutdowns, for example, requires a simulation module. Can be used for preventive maintenance to allow measurements to be released.

If the simulation of the first principles is run concurrently with the process performed by the semiconductor tool, the simulation of the first principles may provide characterization data for the same or different parameters that are tested by the experimental process performed by the simulation. have. For example, the simulation of the first principles can be performed to provide changes to the parameter tested by the design of the experimental process performed by the semiconductor processing tool. Alternatively, the simulation of the first principles may provide characterization data for a parameter different from the parameter tested in an attempt to be performed in the semiconductor processing tool.

Once the simulation is run in step 505, the simulation results are used as part of the data set to characterize the process performed by the semiconductor processing tool, as shown in step 507. As mentioned above, this use as characterization data of simulation results greatly reduces or eliminates the need for time-consuming and expensive experiments required for the design of an experimental approach to characterizing a processor. The characterization data set can be stored in a library for use in subsequent processes performed by the tool.

In addition, the on-tool simulation function of the first principles of the present invention can also be used to provide error detection and process control. Most existing methods for error detection and process control for processes performed by semiconductor processing tools are virtually stochastic. These methods require an experimental design method that involves the burden of performing multiple process executions while changing all of the operating parameters of the tool. The results of these process executions are recorded in a database used for lookup, interpolation, extrapolation, sensitivity analysis, etc. to detect or control the process of the semiconductor processing tool.

However, in order for these probabilistic methods to be able to safely detect and control the tool under widely varying operating conditions, the database must be wide enough to cover all operating conditions, which creates a burden. Becomes The on-tool simulation function of the first principles of the present invention does not require the creation of such a database at all, and given the correct working models and the correct input data, the tool response to process conditions is directly from the physical first principles. And is predicted precisely. However, as more run-time information is available for different operating conditions, statistical methods may still be used to improve the work models and input data, but having this information is for process detection and control functions. It is not required by the present invention. Indeed, the process model can provide a basis by which a process can be empirically controlled by extending such known empirical solutions into "solutions" in which empirical results are not physically formed. Thus, one embodiment of the present invention empirically characterizes a process tool by complementing known (ie, physically observed) solutions with simulation module solutions of the first principles, which simulation module solutions contradict known solutions. It doesn't work. Ultimately, as better statistics are developed, simulation module solutions can be replaced by a database of empirical solutions.

In one embodiment of the present invention, on-tool simulation of the first principles does not require the creation of a database or access to the database, since the tool response to process conditions is directly predicted from the first principles. As more execution time information is available for different operating conditions, statistical methods may still be used to improve the work models and input data, but having this information is useful for process detection and control and error correction. It is not required in the examples.

6 is a block diagram of a system for controlling a process performed by a semiconductor processing tool using simulation techniques of the first principles, in accordance with an embodiment of the present invention. As can be seen in this figure, the system includes a process tool 602 coupled to an advanced process control (APC) infrastructure 604, where the APC infrastructure 604 includes a simulation module 606, APC A controller 608, and a library 610. Also coupled to the APC infrastructure 604 is a metrology tool 612 and a remote controller 614. As can be seen in FIG. 6, the library 610 may include a solution database 616 and a grid database 618.

Process tool 602 may be implemented as semiconductor processing tool 102 described with reference to FIG. 1. Thus, the process tool 602 is, for example, a material processing system, an etching system, a photosensitive resin film spin coating system, a lithography system, an insulation coating system, a deposition system system, a rapid thermal processing (RTP) system for thermal annealing, And / or a batch diffusion furnace or other suitable semiconductor fabrication processing system. As can be seen in FIG. 6, the process tool 602 provides data to the simulation module 606 and receives control data from the APC controller 608, as further described below. Process tool 602 is also coupled to metrology tool 612 that provides information on process results to simulation module 606.

The simulation module 606 is another processing device capable of executing simulation techniques of the first principles to control the process performed by the computer, workstation, or the tool 602, and thus the simulation module described with reference to FIG. And may be implemented as 302. Accordingly, the simulation module 602 executes simulations of the first principles as well as the physical model 106 of the first principles and the simulation processor 108 of the first principles described with reference to FIG. 1 to help control the process. Any other hardware and / or software that may be. In the embodiment of FIG. 6, the simulation module 606 is configured to receive tool data for processing and subsequent use during simulation model execution from one or more diagnostics in the tool 602. The instrument data may include mechanical data, electrical data, chemical data, thermodynamic data of the fluid described above, or any input data described above with reference to FIGS. 1 and 2. In the embodiment of FIG. 6, the tool data may be used to determine boundary conditions and initial conditions for the model to be executed in the simulation module 606. The model can be used to calculate flow fields, electro-magnetic fields, temperature fields, chemical properties, chemical properties of the surface (i.e., chemical properties of the etching surface or deposition surface chemicals), for example, It may include the ANSYS, FLUENT, or CFD-ACE + codes. Models developed from the first principles can determine details within the processing system to provide input for process control of the tool.

The APC controller 608 receives the simulation results from the simulation module 606 and uses the simulation results to implement a control method for process adjustment / correction of the processes performed in the tool 602. It is coupled. For example, the adjustment may be performed to correct for nonuniformities in the processor. In one embodiment of the present invention, one or more perturbation solutions are executed in the simulation module 606, focusing on the process solutions for the process currently executed in the process tool 602. Perturbation solutions are then used, for example, to determine the direction in n-dimensional space for applying correction, for example, the method of deepest decent (SD) methods (Numerical Methods, Dahlquist & Bjorck, Prentice-Hall, Inc., Englewood). Cliffs, NJ, 1974, p. 441; Numerial Recipes, Press et al., Cambridge University Press, Cambridge, 1989, pp. 289-306). Corrections may then be implemented on process tool 602 by APC controller 608. For example, one or more of the tool data (ie physical sensor data) or results from the current run of the simulation shows that the processing system exhibits a non-uniform static pressure field covering the substrate given the current initial / boundary conditions. Can direct. Conversely, this non-uniformity may contribute to the non-uniformity observed by the metrology tool with respect to the metrics used to quantify the performance of the substrate process measured for the substrate, ie critical dimensions, feature depths, film thickness, and the like. By perturbing the input parameters into the current run of the simulation, a set of perturbation solutions can be obtained to determine the best "path" to be taken to eliminate or reduce the static pressure unevenness. For example, input parameters for the process may include pressure, power (delivered to the electrode to generate plasma), gas flow rate, and the like. By perturbing one input parameter at a time and keeping all other input parameters constant, a sensitivity matrix can be formed that can be used in the optimization scheme identified above to derive a correction value suitable for correcting process unevenness. Can be.

In another embodiment of the present invention, simulation results are described in pending US patent application Ser. No. 60/343174, entitled "Method of detecting, identifying, and correcting process performance", the contents of which are incorporated herein by reference. It is used with a principal components analysis (PCA) model that is formulated as follows. There, using multivariate analysis (ie, PCA), the relationship between the simulated signature (ie, spatial components of the simulation model results) and one or more controllable process parameter sets can be determined. This relationship can be used to improve the data profile corresponding to the process performance parameters (ie, model results). Analysis of the principle components determines the relationship between the spatial components of the result (or predicted output) for the simulation of the semiconductor processing tool and the set of one or more control variables (or input parameters). The determined relationship is used to determine corrections to one or more control variables (or input parameters) to minimize the size of the spatial component to improve (or reduce) the non-uniformity of the simulated results (or measured results if available). do.

As noted above, the library 610 coupled to the simulation module 606 of FIG. 6 is configured to include a solution database 616 and a grid database 618. The solution database 616 may comprise an approximate n-dimensional database of solutions, whereby the order n of the n-dimensional space is governed by the number of independent parameters for a given solution algorithm. When the simulation module 606 retrieves tool data for executing a given process, the library 610 may be searched to determine the best solution based on model input. This solution can be used as an initial condition for the simulation of subsequent first principles according to the invention, thereby reducing the number of iterations that must be performed by the simulation module to provide a simulation result. With each model run, a new solution can be added to the solution database 616. Additionally, grid database 618 includes one or more grid sets, whereby each grid set can address a given process tool or process tool geometry. Each grid set may include one or more grids with different grid solutions, ranging from coarse to fine. The selection of grids can be achieved by performing multi-grid techniques (i.e. solving the simulation results on the approximate grid, entailing solutions on a finer grid, involving solutions on the finest grid, etc.) It can be used to reduce.

The metrology tool 612 includes an etch rate, a deposition rate, an etch selectivity (ratio of the rate at which the first material is etched relative to the rate at which the second material is etched), an etching threshold dimension (eg, a length or width of the feature ), Anisotropy of etching features (eg, sidewall profiles of etching features), film properties (eg, film stress, porosity, etc.), film thickness of masks (eg, photosensitive resin films), masks (eg For example, it may be configured to measure process performance parameters such as the pattern critical dimension of the photosensitive resin film (or any other parameter for the process performed by the semiconductor processing tool. The remote controller 612 may be configured with model solver parameters ( Exchange information with the simulation module 606, including model solver parameters (ie, solver parameter updates), solution status, model solutions, and solution convergence history.

7 is a flowchart illustrating a process for controlling a process performed by a semiconductor processing tool using simulation techniques of the first principles, in accordance with an embodiment of the present invention. The flow chart is shown beginning with step 702 for processing a substrate or batch of substrates within a process tool, such as process tool 602. In step 704, the tool data is measured as input and provided to a simulation module, such as simulation module 606. Next, as shown in step 706, boundary conditions and initial conditions for establishing the model are imposed on the physical model of the first principles for the simulation module. In step 708, a physical model of the first principles is executed to provide simulation results of the first principles output to a controller, such as APC controller 608 of FIG. 6. Next, as shown in step 710, the controller determines the control signal from the simulation result. For example, at any time between executions or batches, the operator has the opportunity to select a control algorithm to be used within the APC controller 608. For example, the APC controller can use process model perturbation results or PCA model results. Between executions or batches, the process may be adjusted / corrected by the controller using the simulation results, as shown in step 712.

In another embodiment of the present invention, an empirical model may be used in conjunction with the simulation of the first principles to provide control of the process performed by the process tool. 8 is a block diagram of a system for controlling a process performed by a semiconductor processing tool using simulation techniques and empirical models of first principles, in accordance with an embodiment of the present invention. As can be seen in this figure, the system includes a process tool 802 coupled to an APC infrastructure 804 that includes a simulation module 806 and an APC controller 808. APC infrastructure 804 is also coupled to metrology tool 812 and remote controller 814. These items are similar to the corresponding items discussed with reference to FIG. 6 except that the items of FIG. 8 are additionally configured to operate in consideration of the empirical model. Thus, these similar items are not described with respect to FIG. 8.

As can be seen in FIG. 8, the system includes a model analysis processor 840 coupled to the simulation module 806 and configured to receive simulation results from the module 806. In the embodiment of FIG. 8, model analysis includes empirical model construction from non-dimensionalization of the simulation results. As simulation results are received on a run-to-batch basis, an experience model is constructed and stored in the experience model library 842. For example, process tool 802 experiences a history of process cycles from process development through yield ramp to mass production. During these process cycles, the process chamber of the tool proceeds from the "claen" chamber through chamber qualification and seasoning to the "aged" chamber preceding chamber cleaning and maintenance. . After several maintenance cycles, the empirical model can evolve to include a statistically sufficient sample for the parameter space corresponding to a particular process tool and its associated process. In other words, through cleaning cycles, process cycles, and maintenance cycles, the tool 802 essentially determines the boundaries of the parameter space (with the help of the simulation module). Ultimately, the evolved empirical model stored in the library 842 may replace a generally more process intensive model based on the simulation of the first principles, providing input to the APC controller for process tuning / correction. Can be.

As can be seen in FIG. 8, the remote controller 814 can be used to monitor the evolution of the experience model, and to perform decisions to override the simulation module controller input and select the experience model controller input. 842 may be coupled. In addition, the metrology tool 814 can likewise be coupled to the experience model database to provide input for calibration to the experience model database (connections not shown).

9 is a flowchart illustrating a process for controlling a process performed by a semiconductor processing tool, using simulation techniques and empirical models of the first principles, in accordance with an embodiment of the present invention. The flow chart is shown beginning with step 902 for processing a substrate or batch of substrates within a process tool, such as process tool 802. In step 904, the tool data is measured and provided as input to a simulation module, such as simulation module 806. Next, as shown in step 906, boundary conditions and initial conditions are imposed on the physical model of the first principles for the simulation module to set up the model. In step 908, a physical model of the first principles is executed to perform simulation results of the first principles output for analysis and construction of the empirical model, as shown in step 910.

For example, at any time between executions or batches, the operator has the opportunity to choose process control based on a simulation or empirical model of the first principles. At some point in the empirical model building, the operator prefers an empirical model that can quickly extract controller input for a given set of tool data using a library of data and interpolation / extrapolation schemes at that point, thereby simulating the first principles. You can also choose to override them all. Thus, decision block 912 determines whether to use a simulation of the first principles or an empirical model to control the process. If no override is determined at step 912, the process continues at step 914 where the APC controller determines the control signal from the simulation result. If model override is selected, the APC controller determines a control signal from the empirical model, as shown in step 916. In another embodiment, a combination of simulation results and empirical modeling of the first principles may be used to control the process by the APC controller. As indicated by step 918, the process may be adjusted / corrected by the controller using the model output shown in step 914 or the empirical model output shown in step 916. Thus, the process of FIG. 9 represents a method of in-situ construction of the experience model, and once statistically significant, the experience model may override a computationally intensive simulation process model. During process control, a filter, such as an exponentially weighted moving average (EWMA) filter, may be used to impart only a portion of the required correction. For example, the application of a filter can take the form X _new = (1-λ) X _old + λ (X _predicted -X _old ), where X _new is for a given input parameter (control variable). Is a new value, X _old is a preceding (or previously used) value for a given input parameter, X _predjcted is an expected value for the input parameter based on one of the techniques described above, and λ is between 0 and 1 Is the filter coefficient.

In another embodiment of the present invention, an error detector / classifier may be used in conjunction with the simulation of the first principles to control the process performed by the process tool. 10 is a block diagram of a system for controlling a process performed by a semiconductor processing tool, using the error detectors and simulation techniques of the first principles, in accordance with an embodiment of the present invention. As can be seen in the figure, the system includes a process tool 1002 coupled to an APC infrastructure 1004 that includes a simulation module 1006, an APC controller 1008, and a library 1010. Although not shown in FIG. 10, library 1010 includes a solution database and a grid database. A metrology tool 1012 and a remote controller 1014 are coupled to the APC infrastructure 1004. These items are similar to the corresponding items discussed in connection with FIG. 6, except that the items of FIG. 10 are additionally configured to operate in consideration of error detection. Accordingly, these similar items are not described with reference to FIG. 10.

)를 시뮬레이션 결과들 (Ysim)과 실제 결과들(Yreal)간의 차이를 설명하는 프로세스 성능 데이터(

)와 관련짓는 한 세트의 로딩(또는 상관) 계수들이 정의될 수 있다. As can be seen in FIG. 10, the system includes an error detector 1040 coupled to the simulation module 1006 and configured to receive simulation results from the module 1006. For example, the output of simulation module 1006 can include a profile of data. Next, the profile of the data may serve as an input for multivariate analysis, such as partial least squares (PLS) performed in the error detection apparatus 1040. In PLS analysis, tool perturbation data (

Process performance data describing the difference between simulation results (Ysim) and actual results (Yreal).

) A set of loading (or correlation) coefficients can be defined.

For example, using PLS, observation sets of tool perturbation data are received from the simulation module by the error detector 1040. The instrument perturbation data is determined in the field around the current model solution or in advance in the n-dimensional solution space using the process model. The order n of the n-dimensional parameter space relates to the number of independent parameters of the solution space (ie pressures, mass flow rates, temperatures, etc .; see next).

내에 저장된다. 각각의 관측 세트에 대해, 도구 섭동 데이터는 행렬

의 컬럼으로서 저장될 수 있고, 프로세스 성능 데이터(즉, Ysim-Yreal)는 행렬

의 컬럼으로서 저장될 수 있다. 따라서, 일단 행렬

가 구성되고 나면, 각각의 로우는 상이한 섭동 관측을 표현하고 각각의 컬럼은 상이한 도구 데이터 파라미터를 표현한 다. 일단 행렬

가 구성되고 나면, 각각의 로우는 상이한 관측을 표현하고 각각의 컬럼은 상이한 프로세스 성능 파라미터를 표현한다. 일반적으로, 행렬

는 m×n 행렬일 수 있고, 행렬

는 m×p 행렬일 수 있다. 일단 모든 데이터가 행렬들로 저장되고 나면, 데이터는, 원한다면, 평균-중심화(mean-centered) 및/또는 정규화될 수 있다. 행렬의 컬럼에 저장된 데이터를 평균-중심화하는 프로세스는 컬럼 요소들의 평균값을 계산하고 각 요소에서 평균값을 감산하는 단계를 수반한다. 또한, 행렬의 컬럼에 위치하는 데이터는 컬럼 데이터의 표준 편차에 의해 정규화될 수 있다. For a given set of perturbations, the individual perturbation derivatives (ie, ∂Y / ∂v1, ∂Y / ∂v2, ∂Y / ∂v3; where v1, v2, v3 are different independent parameters) are matrices

Stored within. For each set of observations, the instrument perturbation data is a matrix

And process performance data (ie, Ysim-Yreal) is a matrix

It can be stored as a column of. Thus, once the matrix

Once is constructed, each row represents a different perturbation observation and each column represents a different instrument data parameter. Matrix

Once is configured, each row represents a different observation and each column represents a different process performance parameter. In general, the matrix

Can be an m × n matrix,

May be an m × p matrix. Once all the data has been stored in matrices, the data can be mean-centered and / or normalized if desired. The process of averaging the data stored in the columns of the matrix involves calculating the mean value of the column elements and subtracting the mean value from each element. In addition, the data located in the columns of the matrix can be normalized by the standard deviation of the column data.

In general, in the case of multivariate analysis, the relationship between tool data and process performance data may be expressed as Equation 1 below.

는 상술된 m×n 행렬을 표현하고,

는 n×p(p < n)의 로딩(또는 상관) 행렬을 표현하며,

는 상술된 m×p 행렬을 표현한다. 일단 데이터 행렬들(

및

)이 구성되고 나면, PLS 분석을 사용해,

및

공간들을 최선으로 근사하며

와

간의 상관을 최대화하도록 설계된 관계가 확립된다. From here,

Represents the m × n matrix described above,

Represents a loading (or correlation) matrix of n × p (p <n),

Represents the m × p matrix described above. Once the data matrices (

And

Once configured, use PLS analysis,

And

Best looking at the spaces

Wow

Relationships designed to maximize correlation between them are established.

및

)은 다음과 같이 분해되는데, In the PLS analysis model, matrices (

And

) Is decomposed as follows:

는

변수들을 요약하는 스코어들의 행렬이고,

는 행렬

에 대한 로딩들의 행렬이며,

는

변수들을 요약하는 스코어들의 행 렬이고,

는

와

(

)간의 상관을 표현하는 가중치들의 행렬이며,

,

및

는 나머지들의 행렬들이다. 또한, PLS 분석 모델에서는,

와

를 상관짓는 가중치들이라고 하는 추가 로딩들의

가 존재하는데, 이것은

를 계산하는데 사용된다. From here,

Is

Is a matrix of scores summarizing the variables,

Is a matrix

Is a matrix of loadings for,

Is

Is a matrix of scores summarizing the variables,

Is

Wow

(

Is a matrix of weights representing the correlation between

,

And

Are the matrices of the remainders. In the PLS analysis model,

Wow

Of additional loadings called weights that correlate

Exists, which is

Used to calculate

및

)을 근접하게 근사하며 초평면들(hyper planes)상의 관측 위치들간의 공분산을 최대화할 목적으로, 다차원 공간에서 점들로서 표현되는

및

데이터 모두에 라인, 평면 또는 초평면을 피팅하는 것에 대응된다. In summary, the PLS analysis is geometrically based on the original data tables (

And

) Are represented as points in multidimensional space, with the aim of maximizing covariance between observation positions on hyperplanes.

And

Corresponds to fitting lines, planes or hyperplanes to all of the data.

및

) 및 대응되는 출력들(

,

,

,

,

,

,

,

)과 VIP(variable importance in the projection)에 대한 개략적인 표현을 제공한다. PLS 분석 모델 링을 지원하는 상용 소프트웨어의 일례로는 (The Mathworks, Inc.,Natick, MA로부터 구입 가능한) MATLAB과 함께 제공되는 PLS_Toolbox 또는 (Umetrics, Kinnelon, NJ로부터 구입 가능한) SIMCA-P 8.0을 들 수 있다. 예를 들어, 이러한 소프트웨어에 대한 추가적 세부 사항들은 SIMCA-P 8.0에 대한 User's Manual User Guide에서 제공된다. 다변수 데이터 분석에서의 새로운 표준은 Umetrics AB, Version 8.0인데, 이 또한, 본 발명을 위해 적합하다. 일단 행렬이 공식화되고 나면, 각각의 시뮬레이션 결과에 대해 행렬

가 판정된다. 시뮬레이션된 결과와 실제 결과간에 어떤 차이도, PLA 분석 및 VIP 결과를 사용하여, 판정될 수 있고 특정한 (독립적인) 프로세스 파라미터에 반영될 수 있다. 예를 들어, PLS 모델로부터 출력되는 최대 VIP 값은 차이에 대한 가장 그럴듯한 원인이 될 수 있는 프로세스 파라미터에 해당된다. 11 shows data inputs into PLS analysis (

And

) And corresponding outputs (

,

,

,

,

,

,

,

) And a summary of variable importance in the projection (VIP). One example of commercial software that supports PLS analysis modeling is PLS_Toolbox (available from The Mathworks, Inc., Natick, MA) or SIMCA-P 8.0 (available from Umetrics, Kinnelon, NJ) supplied with MATLAB. Can be. For example, additional details about this software are provided in the User's Manual User Guide for SIMCA-P 8.0. A new standard in multivariate data analysis is Umetrics AB, Version 8.0, which is also suitable for the present invention. Once the matrix is formulated, the matrix for each simulation result

Is determined. Any difference between the simulated and actual results can be determined using the PLA analysis and VIP results and reflected in certain (independent) process parameters. For example, the maximum VIP value output from the PLS model corresponds to a process parameter that may be the most likely cause of the difference.

이 판정된다(그리고, 어쩌면, 라이브러리(1010)에 저장된다). 도구 섭동 데이터(민감도 행렬)가 현재의 모델 솔루션을 중심으로 현장에서 판정되거나, 프로세스 모델을 사용해 n-차원의 솔루션 공간내에서 미리 판정된다. 민감도 행렬 및 실제 결과와 시뮬레이션된 결과간의 차이를 사용하면, 실제 결과와 시뮬레이션된 결과간의 관측된 차이와 최대 상관을 나타내는 제어 변수들(입력 파라미터들)을 식별하기 위해, PLS 분석을 사용해 수학식 1을 풀어낼 수 있다. 상기 예를 사용하면, 프로세스 성능은 기판을 덮고 있는 공간에 대한 정압 프로파일로써 요약될 수 있다. 실제 결과 Yreal은 압력의 측정된 프로파일을 표현하고, Ysim은 압력의 시뮬레이션된 프로파일을 표현한다. 그러나, 가스 유속(gas flow rate)이 정해져 있다고 가정하면, 질량 유량 컨트롤러(mass flow controller)는 유속을 배가한다(그러나, 설정된 값을 보고한다). 압력의 시뮬레이션 프로파일과 측정된(실제) 프로파일간의 차이, 즉, 유속이 실제와 시뮬레이션 경우들 사이에서 2의 팩터만큼 감소된다는 것을 알 수 있을 것이다. 실제와 시뮬레이션 결과들간의 차이는 소정 임계치를 초과하기에 충분할 정도로 클 것이다. PLS 분석을 사용하면, 가스 유속과 같은, 압력의 프로파일에 최대의 영향을 미치는 파라미터들이 식별될 것이다. 오류의 존재 및 그것의 특징은 프로세스 도구 오류 상태로서 오퍼레이터에게 보고되거나, APC 컨트롤러로 하여금 오류 검출에 응답하여 (셧다운과 같은) 프로세스 도구의 제어를 수행하게 할 수 있다. 12 is a flowchart illustrating a process for detecting errors and controlling a process performed by a semiconductor processing tool, using simulation techniques of the first principles, in accordance with an embodiment of the present invention. The flow chart is shown beginning with step 1202 for processing a substrate or batch of substrates within a process tool, such as process tool 1002. In step 1204, the tool data is measured and provided as input to a simulation module, such as simulation module 1006. Next, as indicated by step 1206, boundary conditions and initial conditions for the physical model of the simulation module are imposed to establish the model. In step 1208, a physical model of the first principles is executed to perform simulation results of the first principles output to a controller, such as APC controller 1008 of FIG. 10. For example, at any time between executions or batches, the operator has the opportunity to select a control model to be used within the APC controller. For example, the APC controller can use process model perturbation results or PCA model results. Between executions or batches, the process can be tuned / corrected by the controller using model output. In step 1010, the process model output serves as an input for the PLS model of the error detector 1040, allowing the error to be detected and classified in step 1012. For example, as described above, the difference between actual process performance (Yreal) and simulated (or predicted) process performance (Ysim) for a given process condition (i.e., a set of input control variables) may indicate the presence of a process error. It can be used to determine, where Yreal is measured using a physical sensor or metrology tool, and Ysim is determined by running a simulation provided for the current process condition input. If the difference (or deviation, root mean square, or other statistic) between the actual and simulated results exceeds a certain threshold, it can be predicted that an error has occurred. The predetermined threshold may be, for example, a ratio of the mean value for the particular data (ie, 5%, 10%, 15%) or a multiple of the mean square root of the data (ie, 1σ, 2σ, 3σ). Once an error is detected, the error can be classified using PLS analysis. For example, a sensitivity matrix for a given input condition (ie a set of input control variables)

This is determined (and, perhaps, stored in the library 1010). Tool perturbation data (sensitivity matrices) are determined in the field around the current model solution, or pre-determined in the n-dimensional solution space using the process model. Using the sensitivity matrix and the difference between the actual and simulated results, Equation 1 can be used using PLS analysis to identify control variables (input parameters) that exhibit the maximum correlation with the observed difference between the actual and simulated results. Can be solved. Using the above example, the process performance can be summarized as a static pressure profile for the space covering the substrate. Actual Results Yreal represents the measured profile of pressure, and Ysim represents the simulated profile of pressure. However, assuming a gas flow rate is determined, the mass flow controller doubles the flow rate (but reports the set value). It will be appreciated that the difference between the simulated profile of the pressure and the measured (actual) profile, that is, the flow rate, is reduced by a factor of two between the actual and simulation cases. The difference between the actual and simulation results will be large enough to exceed a certain threshold. Using PLS analysis, the parameters that have the greatest impact on the profile of pressure, such as gas flow rate, will be identified. The presence of the error and its characteristics may be reported to the operator as a process tool error condition, or may cause the APC controller to perform control of the process tool (such as shutdown) in response to the error detection.

13 is a block diagram of a vacuum processing system to which a process control embodiment of the present invention may be applied. The vacuum processing system shown in FIG. 13 is provided for illustrative purposes and does not limit the scope of the invention in any way. The vacuum processing system includes a process tool 1302 with a substrate holder 1304, a gas injection system 1306, and a vacuum pumping system for supporting a substrate 1305. Gas injection system 1306 may include a gas injection plate, a gas injection plenum, and one or more gas injection baffle plates in the gas injection plenum. The gas injection plenum may be coupled to one or more gas sources, such as gas A and gas B, wherein the mass flow rates of gas A and gas B to the processing system are two mass flow controllers (MFCA 1308). And MFCB 1310). In addition, a pressure sensor 1312 for measuring the pressure P1 may be coupled to the gas injection plenum. The substrate holder includes, for example, a helium gas source, an electrostatic clamping system, cooling elements and heating elements to improve gas-gap thermal conductance between the substrate and the substrate holder. And temperature control elements, lift pins for lifting the substrate to and from the surface of the substrate holder, and may include a plurality of components. In addition, the substrate holder may include a temperature sensor 1314 for measuring the temperature T1 or the substrate temperature of the substrate holder and a temperature sensor 1316 for measuring the refrigerant temperature T3. As described above, helium gas is supplied to the backside of the substrate, where the gas-gap pressure P (He) can be changed at one or more locations. In addition, another pressure sensor 1318 may be coupled to the process tool to measure chamber pressure P2, and another pressure sensor 1322 is coupled inside the vacuum pumping system to measure internal pressure P3. can do.

The diagnostic controller 1324 may be configured to be coupled to each of the sensors described above to provide measurements from those sensors to the simulation module described above. In the example system of FIG. 13, the model executed in the simulation module includes, for example, three components: a thermal component, a gas dynamics component, and a chemical property component. In the first component, the gas-gap pressure field may be determined and then the gas-gap thermal conduction coefficient may be calculated. Then, the temperature field spatially determined for the substrate (and substrate holder), boundary conditions (and internal conditions), such as boundary temperature or boundary heat flux, power accumulated in the resistance heating elements, cooling elements Can be determined by appropriately setting the power to be removed at, heat flux at the substrate surface due to the presence of the plasma, and the like.

In one example of the present invention, ANSYS is used to calculate the temperature field. Using a second component of the process model (ie, a gas dynamics component), the gas pressure field and velocity field can be determined using the surface temperatures calculated in the thermal component and some of the above measurements. For example, mass flow rate and pressure P1 may be used to determine internal conditions, pressure P3 may be used to determine external conditions, and CFD-ACE + may be used to calculate gas pressure and velocity fields. Can be. Using a chemical property model (ie, a third component), the previously calculated speed, pressure, and temperature fields can be used as inputs to the chemical property model, for example, to calculate the etch rate. Depending on the complexity of the geometry of the process tool, each of these model components can run in a time scale within a batch-to-batch process cycle. Any one of these components can be used, for example, to provide spatial uniformity data as input for process control, method research, process characterization, and / or error detection / correction.

From models and process analysis derived in response to changes in effects such as processing conditions and / or reactor aging, an empirical model can be assimilated over time. Thus, when the number of iterations for a reactor, as determined by standard statistical analysis programs, becomes statistically significant, process control evolves to experience-based control for processes that are essentially "iterations" of preceding execution operations. do. However, in accordance with the present invention, process control returns to the function of performing a simulation of the first principles if necessary to accommodate new processes or changes in process geometry.

14 illustrates a computer system 1401 in which embodiments of the present invention may be implemented. The computer system 1401 is used as the simulation processor 108 of the first principles for performing any or all of the functions for the simulation processor of the first principles described above, or as described with reference to FIGS. It can be used as any other apparatus for performing any process step. Computer system 1401 includes a processor 1403 coupled with a bus 1402 for processing information with a bus 1402 or other communication mechanism for conveying information. The computer system 1401 may also be a random access memory (RAM) or other dynamic storage device (e.g., a DRAM (DRAM) coupled to the bus 1402 for storing information and instructions to be executed by the processor 1403). main memory, such as dynamic RAM (SRAM), static RAM (SRAM), and synchronous DRAM (SDRAM). In addition, main memory 1404 may be used to store temporary variables or other intermediate information during execution of instructions by processor 1403. Computer system 1401 may be a read only memory (ROM) 1405 or other static storage device (eg, PROM (programmable) coupled to bus 1402 to store static information and instructions for processor 1403). ROM), erasable PROM (EPROM), and electrically erasable PROM (EEPROM).

The computer system 1401 also includes a magnetic hard disk 1407 and removable media drive 1408 (e.g., floppy disk drive, read-only CD drive, read / write CD drive, CD juke) for storing information and instructions. Also included is a disk controller 1406 coupled to the bus 1402 to control one or more storage devices, such as boxes, tape drives, and removable magneto-optical drives. Storage devices use a suitable device interface (for example, a small computer system interface (SCSI), integrated device electronics (IDE), enhanced-IDE (E-IDE), direct memory access (DMA), or ultra-DMA)). It may be added to the system 1401.

Computer system 1401 may also include special purpose logic devices (eg, application specific integrated circuits (ASICs)) or configurable logic devices (eg, simple programmable logic devices (SPLDs), CPLDs). (complex programmable logic devices), and field programmable gate arrays (FPGAs).

The computer system 1401 also includes a display controller 1409 coupled to the bus 1402 to control the display 1410, such as a cathode ray tube (CRT) for displaying information to a computer user. It may include. The computer system includes input devices, such as keyboard 1411 and pointing device 1412, for interacting with a computer user and providing information to processor 1403. The pointing device 1412 may be, for example, a mouse, trackball, or pointing stick for communicating indication information and command selections to the processor 1403 and controlling cursor movement on the display 1410. The printer may also provide printed lists of data stored and / or generated by the computer system 1401.

Computer system 1401 performs some or all of the processing steps of the present invention in response to processor 1403 executing one or more sequences of one or more instructions contained in the memory, such as main memory 1404. . These instructions may be read into main memory 1404 from another computer readable medium, such as hard disk 1407 or removable media drive 1408. One or more processors in a multi-processing configuration may be used to execute the sequences of instructions contained in main memory 1404. In other embodiments, a wired circuit may be used in place of or in conjunction with software instructions. As such, the embodiments are not limited to any specific combination of hardware circuitry and software.

As noted above, computer system 1401 holds one or more computer readable instructions for holding instructions programmed according to the teachings of the present invention and for containing data structures, tables, records, or other data described herein. Media or memory. Examples of computer readable media include CDs, hard disks, floppy disks, tape, magneto-optical disks, PROMs (EPROM, EEPROM, flash EPROM), DRAM, SRAM, SDRAM, or any other magnetic media, CDs (eg, CD-ROM), or any other optical medium, punched cards, paper tape, or other physical medium with patterns of holes, carrier wave (described below), or computer readable Any other medium.

The present invention is directed to a computer system 1401 for controlling a computer system 1401, for driving an apparatus or devices for implementing the invention, and stored in any one or a combination of computer readable media. It includes software to enable interaction with this human user (e.g., a print production person). Such software may include, but is not limited to, device drivers, operating systems, development tools, and application software. Such computer readable media further comprise the computer program product of the present invention for performing all or part of the processing performed when implementing the invention (if processing is distributed).

Computer code devices of the present invention include, but are not limited to, scripts, interpretable programs, dynamic link libraries, Java classes, and complete executable programs. May be an interpretable or executable code mechanism. In addition, some of the processing of the present invention may be distributed for better performance, reliability, and / or cost.

The term "computer readable medium" as used herein is used to refer to any medium that participates in providing instructions to the processor 1403 for execution. Computer-readable media may include, but are not limited to, non-volatile media, volatile media, and transmission media. Non-volatile media include optical, magnetic disks, and magneto-optical disks, such as, for example, hard disk 1407 or removable media drive 1408. Volatile media include dynamic memory, such as main memory 1404. Transmission media include coaxial cables, copper wire and optical fibers, including the wirings that make up the bus 1402. Transmission media may take the form of acoustic or light waves, such as those generated during radio wave and infrared data communications.

Various forms of computer readable media may be involved in carrying one or more sequences of one or more instructions to a processor 1403 for execution. For example, instructions may initially be delivered to a magnetic disk of a remote computer. The remote computer can remotely load instructions for implementing all or part of the present invention into dynamic memory and send the instructions over a telephone line using a modem. A modem belonging to computer system 1401 may receive data over a telephone line and convert the data into an infrared signal using an infrared transmitter. An infrared detector coupled to the bus 1402 may receive data transferred as an infrared signal and place the data on the bus 1402. Bus 1402 passes data to main memory 1404, where processor 1403 retrieves and executes instructions. Instructions received by main memory 1404 may optionally be stored in storage 1407 or 1408 before or after execution by processor 1403.

Computer system 1401 also includes a communication interface 1413 coupled to bus 1402. The communication interface 1413 provides a bidirectional data communication coupling to a network link 1414 that is connected to another communication network 1416, such as, for example, a local area network (LAN) 1415 or the Internet. For example, communication interface 1413 may be a network interface card for attaching to any packet switched LAN. As another example, communication interface 1413 may be an asymmetrical digital subscriber line (ADSL) card, an integrated services digital network (ISDN) card, or a modem to provide a data communication connection to a corresponding type of communication line. Wireless links may be implemented. In any such implementation, communication interface 1413 transmits and receives electronic, electromagnetic or optical signals that carry digital data streams representing various types of information.

Network link 1414 typically provides data communication to other data devices over one or more networks. For example, the network link 1414 may be connected to another computer via a local network 1415 (eg, a LAN) or through equipment operated by a service provider that provides communication services via the communication network 1416. You can also provide a connection. Local network 1414 and communication network 1416 carry, for example, electrical, electromagnetic, or optical signals and associated physical layers (eg, CAT 5 cables, coaxial cables, optical fibers, etc.) that carry digital data streams. Use Signals passing through the various networks, which pass digital data to and from the computer system 1401, and signals passing through the communication interface 1413 at the network link 1414 may be baseband signals or carrier based. It can be implemented with signals. Baseband signals convey digital data as unmodulated electrical pulses that describe a stream of digital data bits, wherein the term “bits” should be interpreted broadly to mean a symbol, each symbol having at least one It carries the above information bits. Digital data may be used to modulate a carrier wave, such as by using amplitude, phase, and / or frequency shifted signals propagated through or transmitted as electromagnetic waves through the conducting medium. As such, digital data may be transmitted as unmodulated baseband data over a “wired” communication channel and / or within a predetermined frequency band that is different from the baseband by modulating a carrier wave. Computer system 1401 may transmit and receive data including program code via network (s) 1415 and 1416, network link 1414, and communication interface 1413. The network link 1414 may also provide a connection to a mobile device 1417, such as a personal digital assistant (PDA) laptop computer or cellular telephone, via a LAN 1415.

Many modifications and variations of the present invention are possible in light of the above teachings. It is, therefore, to be understood that within the scope of the appended claims, the invention may be practiced otherwise than as specifically described herein. For example, the process steps described herein and referred to in the claims may be performed simultaneously or in a sequence other than the sequence in which the process steps are described or enumerated herein. Those skilled in the art should understand that the process steps necessary for the performance of subsequent process steps need only be performed before the subsequent process steps are performed.

Claims (110)

A method of facilitating a process performed by a semiconductor processing tool, Inputting data related to a process performed by the semiconductor processing tool; Inputting a physical model of first principles related to the semiconductor processing tool; Performing simulations of first principles using the input data and the physical model to provide virtual sensor measurements related to the process performed by the semiconductor processing tool; And Using the virtual sensor measurements to facilitate the process performed by the semiconductor processing tool A method for facilitating a process performed by a semiconductor processing tool having a. The method of claim 1, wherein the inputting comprises directly inputting data related to a process performed by the semiconductor processing tool from at least one of a physical sensor and a metrology tool physically mounted on the semiconductor processing tool. Method. The method of claim 1, wherein the inputting comprises indirectly inputting data related to a process performed by the semiconductor processing tool from one or more of a manual input device and a database. 4. The method of claim 3, wherein the indirectly inputting comprises inputting recorded data from a process previously performed by the semiconductor processing tool. 4. The method of claim 3, wherein the indirectly inputting comprises inputting data set by a simulation operator. 2. The method of claim 1, wherein entering data comprises entering data related to one or more of the semiconductor processing tool and physical features of the semiconductor tool environment. The method of claim 1, wherein entering data comprises entering data related to one or more of the characteristics and results of a process performed by the semiconductor processing tool. The method of claim 1, wherein inputting a physical model of the first principles comprises inputting a spatially determined model for the geometry of the semiconductor processing tool. 2. The method of claim 1, wherein inputting a physical model of the first principles comprises inputting basic equations necessary to perform a simulation of the first principles to obtain a virtual sensor reading. 2. The method of claim 1, wherein performing a simulation of the first principles comprises performing a simulation of the first principles concurrently with the process performed by the semiconductor processing tool. The method of claim 10, Iteratively updating data from the physical sensor or metrology tool during the semiconductor process; Iteratively performing a simulation of the first principles using the updated data during the semiconductor process; And And based on the virtual sensor measurements obtained during the semiconductor process, facilitating the semiconductor process while executing the semiconductor process. The method of claim 10, Establishing boundary conditions for the simulation of the first principles prior to starting the semiconductor process; Performing a time dependent simulation of the semiconductor process during the semiconductor process and without direct input from the semiconductor process; And And based on the virtual sensor measurements obtained during the semiconductor process, facilitating the semiconductor process while executing the semiconductor process. The method of claim 1, wherein performing a simulation of the first principles comprises performing a simulation of the first principles asynchronously with a process performed by the semiconductor processing tool. 14. The method of claim 13, wherein entering data comprises entering one or more of initial and boundary conditions for a simulation of the first principles recorded from a process previously performed. 4. The method of claim 3, wherein the indirectly inputting step comprises inputting best known input parameters for the physical model. The method of claim 15, Comparing the virtual sensor measurements with actual sensor measurements; And Refining one or more of the best known input parameters and the physical model to obtain a better match between the virtual sensor measurements and actual sensor measurements. The method further comprising. 2. The method of claim 1, wherein using the virtual sensor measurements comprises using the virtual sensor measurements to characterize a process performed by the semiconductor processing tool. 4. The method of claim 1, wherein using the virtual sensor measurements comprises using the virtual sensor measurements to control a process performed by the semiconductor processing tool. 2. The method of claim 1, wherein using the virtual sensor measurements comprises using the virtual sensor measurements to detect errors in a process performed by the semiconductor processing tool. The method of claim 1, further comprising storing the virtual sensor measurements in a library for subsequent use in simulation of first principles. The method of claim 1, further comprising using a network of interconnected resources to perform one or more of the process steps described in claim 1. 22. The method of claim 21, further comprising using code parallelization between interconnected computational resources to share the computational load of the simulation of the first principles. 22. The method of claim 21, further comprising sharing simulation information between interconnected resources to facilitate the process performed by the semiconductor processing tool. 24. The method of claim 23, wherein sharing the simulation information comprises: simulating simulation results between the interconnected resources to reduce redundant execution of simulations of substantially similar first principles by different resources. Dispensing. 24. The method of claim 23, wherein sharing the simulation information further comprises: distributing model changes among the interconnected resources to reduce redundant refinement of simulations of the first principles by different resources. And a step of making it. 22. The method of claim 21, further comprising using remote resources over a wide area network (WAN) to facilitate a semiconductor process performed by the semiconductor processing tool. 27. The method of claim 26, wherein using remote resources comprises using one or more of remote computing and storage resources over a WAN to facilitate a semiconductor process performed by the semiconductor processing tool. How to do. A semiconductor processing tool configured to perform the process; An input device configured to input data related to a process performed by the semiconductor processing tool; And As a simulation processor of the first principles, Input a physical model of first principles related to the semiconductor processing tool; Using the input data and the physical model to provide the virtual sensor measurements related to the process performed by the semiconductor processing tool, which are used to facilitate the process performed by the semiconductor processing tool. A simulation processor of the first principles, configured to perform a simulation of the first principles A semiconductor processing system comprising a. 29. The system of claim 28, wherein the input device includes one or more of a physical sensor and a metrology tool physically mounted on the semiconductor processing tool. 29. The system of claim 28, wherein the input device comprises one or more of a manual input device and a database. 31. The system of claim 30, wherein the input device is configured to input data recorded from a process previously performed by the semiconductor processing tool. 31. The system of claim 30, wherein the input device is configured to input data set by a simulation operator. 29. The system of claim 28, wherein the input device is configured to input data related to one or more of the physical characteristics of the semiconductor processing tool and the semiconductor tool environment. 29. The system of claim 28, wherein the input device is configured to input data related to one or more of the characteristics and results of a process performed by the semiconductor processing tool. 29. The system of claim 28, wherein the processor is configured to input a physical model of first principles having a spatially determined model of the geometry of the semiconductor processing tool. 29. The system of claim 28, wherein the processor is configured to input a physical model of first principles having basic equations needed to perform a simulation of the first principles to obtain a virtual sensor reading. 29. The system of claim 28, wherein the processor is configured to perform the simulation of the first principles simultaneously with the process performed by the semiconductor processing tool. The processor of claim 37, wherein the processor is configured to: Iteratively updating data from the physical sensor or metrology tool during the semiconductor process; Further configured to repeatedly perform the simulation of the first principles using the updated data during the semiconductor process, The semiconductor process is facilitated simultaneously with executing the semiconductor process based on virtual sensor measurements obtained during the semiconductor process. The processor of claim 37, wherein the processor is configured to: Establish boundary conditions for the simulation of the first principles prior to starting the semiconductor process; Is further configured to perform a time dependent simulation of the semiconductor process during the semiconductor process and without direct input from the semiconductor process, The semiconductor process is facilitated simultaneously with executing the semiconductor process based on virtual sensor measurements obtained during the semiconductor process. 29. The system of claim 28, wherein the processor is configured to perform the simulation of the first principles asynchronously with the process performed by the semiconductor processing tool. 41. The first principle of claim 40, wherein the processor uses the input data to set at least one of an initial and boundary conditions of a simulation of the first principles recorded from a previously performed process. The system being configured to perform a simulation. 31. The system of claim 30, wherein the input device is configured to input the best known input parameters for the physical model. The processor of claim 42, wherein the processor is configured to: Compare the virtual sensor measurements with actual sensor measurements; And to improve one or more of the best known input parameters and the physical model to obtain a better match between the virtual sensor measurements and the actual sensor measurements. 29. The system of claim 28, wherein the virtual sensor measurements are used to characterize a process performed by the semiconductor processing tool. 29. The system of claim 28, wherein the virtual sensor measurements are used to control a process performed by the semiconductor processing tool. 29. The system of claim 28, wherein the virtual sensor measurements are used to detect errors in a process performed by the semiconductor processing tool. 29. The system of claim 28, wherein the processor is further configured to store the virtual sensor measurements in a library for subsequent use in the simulation of the first principles. 29. The system of claim 28, connected to the processor, the processor being configured to assist in performing one or more of inputting a simulation model of first principles and performing a simulation of first principles. Further comprising a network of connected resources. 49. The system of claim 48, wherein the network of interconnected resources is configured to use code parallelism with the processor to share the computational load of the simulation of the first principles. 49. The system of claim 48, wherein the network of interconnected resources is configured to share simulation information with the processor to facilitate the process performed by the semiconductor processing tool. The system of claim 50, wherein the network of interconnected resources is configured to distribute simulation results to the processor to reduce redundant execution of simulations of substantially similar first principles. 51. The system of claim 50, wherein the network of interconnected resources is configured to distribute model changes to the processor to reduce redundant improvements in simulations of the first principles. The system of claim 48, further comprising remote resources connected to the processor via a WAN and configured to facilitate a semiconductor process performed by the semiconductor processing tool. 54. The system of claim 53, wherein the remote resources comprise one or more of computational and storage resources. A system for facilitating a process performed by a semiconductor processing tool, the system comprising: Means for inputting data related to a process performed by the semiconductor processing tool; Means for inputting a physical model of first principles related to the semiconductor processing tool; Means for performing simulations of first principles using the input data and the physical model to provide virtual sensor measurements related to the process performed by the semiconductor processing tool; And Means for using the virtual sensor measurements to facilitate the process performed by the semiconductor processing tool A process facilitating system for facilitating a process having a. 56. The system of claim 55, wherein the means for performing a simulation of the first principles comprises means for performing a simulation of the first principles simultaneously with the process performed by the semiconductor processing tool. The method of claim 56, wherein Means for iteratively updating data from the physical sensor or metrology tool during the semiconductor process; Means for iteratively performing a simulation of the first principles using the updated data during the semiconductor process; And And means for facilitating the semiconductor process concurrently with executing the semiconductor process based on virtual sensor measurements obtained during the semiconductor process. The method of claim 56, wherein Means for setting boundary conditions for the simulation of the first principles prior to starting the semiconductor process; Means for performing a time dependent simulation of the semiconductor process during the semiconductor process and without direct input from the semiconductor process; And And means for facilitating the semiconductor process concurrently with executing the semiconductor process based on virtual sensor measurements obtained during the semiconductor process. 56. The apparatus of claim 55, further comprising means for inputting best known input parameters for the physical model, The system, Means for comparing the virtual sensor measurements with actual sensor measurements; And And means for improving one or more of said best known input parameters and said physical model to obtain a better match between said virtual sensor measurements and actual sensor measurements. 56. The system of claim 55, further comprising means for sharing the computational load of the simulation of the first principles. 42. The system of claim 41, further comprising means for sharing simulation information between interconnected resources to facilitate the process performed by the semiconductor processing tool. A computer recording medium comprising program instructions for execution on a processor, wherein the processor, when executed by a computer system, Inputting data related to a process performed by the semiconductor processing tool; Inputting a physical model of first principles related to the semiconductor processing tool; Performing simulations of first principles using the input data and the physical model to provide virtual sensor measurements related to the process performed by the semiconductor processing tool; And Using the virtual sensor measurements to facilitate the process performed by the semiconductor processing tool Computer-readable media comprising program instructions for causing the computer to perform the command. A method of facilitating a process performed by a semiconductor processing tool, Inputting data related to the process performed by the semiconductor processing tool; Inputting a physical model of first principles related to the semiconductor processing tool; Performing simulations of first principles using the input data and the physical model to provide simulation results for the process performed by the semiconductor processing tool; And Using the simulation results as part of a data set characterizing a process performed by the semiconductor processing tool A method for facilitating a process comprising: 64. The method of claim 63, wherein the inputting step comprises directly inputting data relating to a process performed by the semiconductor processing tool from at least one of a physical sensor and a metrology tool physically mounted on the semiconductor processing tool. It is provided with. 64. The method of claim 63, wherein the inputting comprises indirectly inputting data related to a process performed by the semiconductor processing tool from one or more of a manual input device and a database. 67. The method of claim 65, wherein indirectly inputting comprises inputting recorded data from a process previously performed by the semiconductor processing tool. 66. The method of claim 65, wherein indirectly inputting comprises inputting data set by a simulation operator. 64. The method of claim 63, wherein the inputting comprises inputting data related to a process performed by the semiconductor processing tool as virtual sensor data from a simulation module. 64. The method of claim 63, wherein entering data comprises entering data related to one or more of the semiconductor processing tool and physical features of the semiconductor tool environment. 64. The method of claim 63, wherein entering data comprises entering data related to one or more of the characteristics and results of a process performed by the semiconductor processing tool. 64. The method of claim 63, wherein inputting a physical model of the first principles comprises inputting a spatially determined model of the geometry of the semiconductor processing tool. 64. The method of claim 63, wherein inputting a physical model of the first principles simulates the first principles to obtain a simulation result that can form part of a data set that characterizes a process performed by the semiconductor processing tool. Inputting basic equations necessary to perform the method. 64. The method of claim 63, wherein performing a simulation of the first principles comprises performing a simulation of the first principles concurrently with the process performed by the semiconductor processing tool. 74. The method of claim 73, wherein performing a simulation of the first principles is to modify the parameters of the first principles to provide a simulation result that is a variation of a parameter tested by a concurrent process performed by the semiconductor processing tool. Performing the simulation. 80. The method of claim 73, wherein performing a simulation of the first principles comprises: providing simulation results related to a parameter different from a parameter tested by a concurrent process performed by the semiconductor processing tool. Performing the simulation. 64. The method of claim 63, wherein performing the simulation of the first principles comprises performing a simulation of the first principles asynchronously with the process performed by the semiconductor processing tool. 64. The method of claim 63, further comprising storing the data set in a library for subsequent use in processes performed by the semiconductor processing tool. 64. The method of claim 63, further comprising using a network of interconnected resources to perform one or more of the process steps described in claim 1. 79. The method of claim 78, further comprising using code parallelism between interconnected computational resources to share the computational load of the simulation of the first principles. 80. The method of claim 78, further comprising sharing simulation information between interconnected resources to facilitate the process performed by the semiconductor processing tool. 81. The method of claim 80, wherein sharing the simulation information further comprises: distributing simulation results among the interconnected resources to reduce redundant execution of simulations of substantially similar first principles by different resources. It is provided with. 81. The method of claim 80, wherein sharing the simulation information comprises distributing model changes among the interconnected resources to reduce redundant improvements in simulations of the first principles by different resources. How. The method of claim 80, further comprising using remote resources over a WAN to facilitate a semiconductor process performed by the semiconductor processing tool. 84. The method of claim 83, wherein using remote resources comprises using one or more of remote computing and storage resources over a WAN to facilitate a semiconductor process performed by the semiconductor processing tool. How to do. A semiconductor processing tool configured to perform the process; An input device configured to input data related to a process performed by the semiconductor processing tool; And As a simulation processor of the first principles, Input a physical model of first principles related to the semiconductor processing tool; As a simulation result of first principles for a process performed by the semiconductor processing tool, to provide a simulation result of the first principles used as part of a data set characterizing a process performed by the semiconductor processing tool, And a simulation processor of the first principles, configured to perform a simulation of first principles using the input data and the physical model. 86. The system of claim 85, wherein the input device includes one or more of a physical sensor and a metrology tool physically mounted on the semiconductor processing tool. 86. The system of claim 85, wherein the input device comprises one or more of a manual input device and a database. 88. The system of claim 87, wherein the input device is configured to input data recorded from a process previously performed by the semiconductor processing tool. 88. The system of claim 87, wherein the input device is configured to input data set by a simulation operator. 86. The system of claim 85, wherein the input device is configured to input data related to a process performed by the semiconductor processing tool as virtual sensor data from a simulation module. 86. The system of claim 85, wherein the input device is configured to input data related to one or more of the semiconductor processing tool and physical features of the semiconductor tool environment. 86. The system of claim 85, wherein the input device is configured to input data related to one or more of the characteristics and results of a process performed by the semiconductor processing tool. 86. The system of claim 85, wherein the processor is configured to input a physical model of first principles having a spatially determined model for the geometry of the semiconductor processing tool. 86. The apparatus of claim 85, wherein the processor includes basic equations necessary to perform a simulation of the first principles to obtain a simulation result that can form part of a data set characterizing a process performed by the semiconductor processing tool. And to input a physical model of the first principles. 86. The system of claim 85, wherein the processor is configured to perform the simulation of the first principles concurrently with the process performed by the semiconductor processing tool. 95. The processor of claim 95, wherein the processor is configured to perform simulations of the first principles to provide a simulation result that is a variation of a parameter tested by a concurrent process performed by the semiconductor processing tool. System. 95. The apparatus of claim 95, wherein the processor is configured to perform simulations of the first principles to provide simulation results related to parameters different from the parameters tested by the concurrent process performed by the semiconductor processing tool. System. 86. The system of claim 85, wherein the processor is configured to perform the simulation of the first principles asynchronously with the process performed by the semiconductor processing tool. 86. The system of claim 85, wherein the processor is further configured to store the data set in a library for subsequent use in processes performed by the semiconductor processing tool. 88. The interconnect of claim 85, connected to the processor and configured to support the processor to perform one or more of inputting a simulation model of first principles and performing a simulation of first principles. Further comprising a network of prepared resources. 101. The system of claim 100, wherein the network of interconnected resources is configured to use code parallelization with the processor to share the computational load of the simulation of the first principles. 101. The system of claim 100, wherein the network of interconnected resources is configured to share simulation information with the processor to facilitate the process performed by the semiconductor processing tool. 107. The system of claim 102, wherein the network of interconnected resources is configured to distribute simulation results to the processor to reduce redundant execution of simulations of substantially similar first principles. 107. The system of claim 102, wherein the network of interconnected resources is configured to distribute model changes to the processor to reduce redundant improvements in simulations of first principles. 101. The system of claim 100, further comprising remote resources connected to the processor via a WAN and configured to facilitate a semiconductor process performed by the semiconductor processing tool. 107. The system of claim 105, wherein the remote resources comprise one or more of a computation and storage resource. A system for facilitating a process performed by a semiconductor processing tool, the system comprising: Means for inputting data related to a process performed by the semiconductor processing tool; Means for inputting a physical model of first principles related to the semiconductor processing tool; Means for performing a simulation of first principles using the input data and the physical model to provide a simulation result for the process performed by the semiconductor processing tool; And Means for using the simulation result as part of a data set characterizing a process performed by the semiconductor processing tool A system for facilitating a process comprising a. 107. The system of claim 105, further comprising means for sharing the computational load of the simulation of the first principles. 107. The system of claim 105, further comprising means for sharing simulation information between interconnected resources to facilitate the process performed by the semiconductor processing tool. A computer readable medium containing program instructions for execution on a processor, wherein the processor, when executed by a computer system, Inputting data related to a process performed by the semiconductor processing tool; Inputting a physical model of first principles related to the semiconductor processing tool; Performing simulations of first principles using the input data and the physical model to provide simulation results for the process performed by the semiconductor processing tool; And Using the simulation results as part of a data set characterizing a process performed by the semiconductor processing tool Computer-readable media comprising program instructions for causing the computer to perform the command.

Images